Solar twins in the ELODIE archive

0.3 MB·

by  Dhiaa Mahdi

Astronomy&Astrophysicsmanuscriptno.solar_twins_AA (cid:13)cESO2016
January8,2016
⋆
Solar twins in the ELODIE archive
D.Mahdi1,C.Soubiran1,S.Blanco-Cuaresma2,andL.Chemin1
1 LABUMR5804,Univ.Bordeaux-CNRS,F-33270,Floirac,France
e-mail:[email protected]
2 ObservatoiredeGenève,UniversitédeGenève,CH-1290Versoix,Switzerland
ReceivedJanuary8,2016;acceptedXX
6 ABSTRACT
1
0 Alargedatasetof∼2800spectraextractedfromtheELODIEarchivewasanalysedinordertofindsolartwins.Alistofstellarspectra
2 closelyresemblingthespectrumoftheSunwasselectedbyapplying apurelydifferential method, directlyonthefluxes.Assolar
reference,18spectraofasteroids,Moonandblueskywereused.Atmosphericparametersanddifferentialabundancesof8chemical
n
elementsweredeterminedforthesolartwincandidates,afteracarefulselectionofappropriatelines.TheLifeatureofthetargetswas
a
investigatedandadditionalinformationonabsolutemagnitudeandagewasgatheredfromtheliterature.HIP076114(HD138573)is
J
ourbesttwincandidate,lookingexactlyliketheSuninalltheseproperties.

7
Keywords. Stars:solar-type–Stars:fundamentalparameters–Stars:atmospheres–Stars:abundances

]
R
S1. Introduction analysis of many candidates (CayreldeStrobeletal. 1981;
h. CayreldeStrobel&Bentolila1989;Frieletal.1993),andare-
The definition of solar twin was first introduced by
p view of the state of the art (CayreldeStrobel 1996). More re-
CayreldeStrobeletal. (1981). Solar twins are stars which
- cently, a large scale search of solar twins has been performed
ohave the same physical properties as those of the Sun : mass,
by Datsonetal. (2012, 2014, 2015) using the ESO FEROS and
rradius, luminosity, chemical composition, rotation, activity.
t HARPS archives, with 17 twins identified. PortodeMelloetal.
sConsequently, the spectrum of a solar twin should be identical
(2014) conducted a spectroscopic survey of solar twin stars
atothatoftheSun.Awaytofindsolartwinsisthustocompare
[ within 50 parsecs of the Sun, and identified some candidates
stellar spectra to solar spectra, and to identify those looking on the basis of photometric colours and atmospheric param-

1the most similar. There are several motivations to search for eters determined from high resolution spectra. Several other
vsolar twins. The Sun, as the best known star, is used as the
studies have used the differential analysis of high resolution
9fundamental standard in many astronomical calibrations. A
spectra to identify solar twins but very few stars were found
9motivationtoidentifystarsthatreplicatethesolarastrophysical
very similar to the Sun. HIP079672 (HD146233 - 18 Sco)
5properties is the need to have other reference stars, observable
is the most studied one, and often claimed to be the best
1
duringthe nightunderthe same conditionsas any other target.
0 one(PortodeMello&daSilva 1997; Soubiran&Triaud2004;
For example,Casagrandeetal. (2010) used a set ofsolar twins
. Meléndezetal. 2006; Takeda&Tajitsu 2009), but it was fi-
1tocalibratetheeffectivetemperaturescalefromtheinfraredflux
nally found not to be a perfect twin, with slighlty higher ef-
0method, and showed that this model independent technique is
fective temperature and surface gravity, different chemicalpat-
6not affected by systematics (Casagrandeetal. 2014). Sun-like
tern (Meléndezetal. 2014a), higher mass and younger age
1
stars can also help us to understand whether the chemical
: (Bazotetal. 2011; Lietal. 2012). HIP114328 (HD218544) is
vcomposition of the Sun compared to other stars is unusual consideredbyMeléndezetal.(2014b)asanexcellentsolartwin
i(Ramírezetal. 2009; Nissen 2015) and to explore the connec-
X tohosta rockyplanetduetoits abundancepatternverysimilar
tionsbetweenplanetformationandstellarchemicalcomposition
tothatoftheSun,despitealowerLicontentandindicationsof
r(Ramírezetal.2014b).Solartwinsrepresentnaturalcandidates
a beingolder.TheabundancepatternofHIP102152(HD197027)
for harboring planetary systems similar to ours. Finally solar
wasclaimedtobethe mostsimilar tosolar ofanyknownsolar
twins may also give some clues on where and how the Sun
twinbyMonroeetal.(2013),whileHIP056948(HD101346)is
formed in our Galaxy. Such stars may have formed from the
consideredasaprimetargetinthequestforotherEarthsdueto
sameconditionsastheSunandfromthesamemolecularcloud.
itssimilaritytotheSun(Meléndezetal.2012;Takeda&Tajitsu
In that case, one refers to solar siblings which are expected to
2009; Meléndez&Ramírez 2007). In fact, most of the solar
alsosharethesamekinematicalpropertiesastheSun(Liuetal.
twins are found to have chemical compositions different from
2015).
that of the Sun when submitted to high precision differential
Since the pioneering work by Hardorp (1978), the hunt
analysis,suggestingthattheSunhasanunusualabundancepat-
of the closest solar twins has been very active. Giusa
tern (Nissen 2015). Interestingly Önehagetal. (2011) claimed
Cayrel de Strobel and her collaborators made a huge con-
thatM67-1194isthefirstsolartwinknowntobelongtoastellar
tribution to the subject with the detailed spectroscopic
association.ThechemicalsimilaritybetweentheSunandM67-
1194 suggests that the Sun once formed in a cluster like M67.
⋆ BasedondataextractedfromtheELODIEarchiveatObservatoire
HoweverPichardoetal.(2012)demonstratedwithdynamicalar-
deHaute–Provence(OHP),http://atlas.obs-hp.fr/elodie/
Articlenumber,page1of15
A&Aproofs:manuscriptno.solar_twins_AA
guments that M67 could not have been the birth cluster of the done.Largeseriesofobservationsofasinglestarwereshortened
SunandalsoexcludedthepossibilitythattheSunandM67were byretrievingfromthearchiveonlyoneobservationpernightand
borninthesamemolecularcloud. fiveobservationsintotal(theoneswiththehighestS/N,favour-
Inthispaperwesearchedfornewsolartwinsinalargesam- ing exposures without simultaneous thorium-argon). Repeated
pleofspectraselectedfromtheELODIEarchive(Moultakaetal. observationsofthesamestarswereusedtotesttheconsistency
2004), using 18 spectra of asteroids, the Moon and the day ofourmethodsandtoprovidemorerobustresults,asexplained
sky as solar reference.This observationalmaterial is described inSect.4andSect.6.Theselectioncorrespondsto2784spec-
in Sect. 2. We proceeded in several steps. First, as described traof1165differentstarswhichweredownloadedasFITSfiles
in Sect. 3, twin candidates were identified with a minimum in S2D and spec formats, together with the CCF files. In addi-
distance method already used and validated for that task by tion, 18 spectra of solar system bodies and day sky light were
Soubiran&Triaud (2004). Then their atmospheric parameters retrievedtobeusedasreferencesolarspectra.Thesesolarspec-
(effective temperature T , surface gravity logg, metallicity tra were obtained at differentdates, with variousS/N and fiber
eff
[M/H]) were derived and compared to those of the Sun (Sect. modes giving a representive range of observing conditions for
4).WeinspectedthespectralrangearoundtheLilinesat670.78 theotherstars.ThelistofsolarspectraispresentedinTable1.
nmtofindtwincandidatesshowingthesameLidepletionasthe
Sun (Sect. 5). In Sect. 6 we explain how we selected the good Table1.SolarspectraavailableintheELODIEarchive.Thespectracan
linesfromwhichabundanceswheredetermineddifferentiallyto easilybe retrievedfromthearchive usingthedateof observation and
the solar spectra. Finally we searched for extra information on the exposure serial number (imanum). S/N is provided in the header
thetargets(Sect.7)anddiscussourfindingsinSect.8. ofthespectrum, fortheorder 46centeredon550 nm.TheFWHMis
thatoftheCCFretrievedfromthearchive.Theexposuretypeindicates
whethertheobservationwasmadewithasimultaneousthoriumcalibra-
2. ELODIEspectra tion(OBTH)ornot(OBJO)andwithdirectorscrambledfibers(dand
srespectively,after1997).
Allthespectrausedinthispaperwereretrievedfromthearchive
of the ELODIE echelle spectrograph (Moultakaetal. 2004). Name Date/Imanum S/N FWHM Exposure
ELODIEwasonthe 1.93mtelescopeatObservatoiredeHaute– 550nm kms−1 type
Provence (OHP) between late 1993 and mid 2006. It was de- CERES 19950206/0021 107.3 11.13 OBTH
signedforverypreciseradialvelocitymeasurements(thediscov- CERES 20000327/0023 117.3 11.01 OBJOd
eryof thefirst extra-solarplanet51PegB byMayor&Queloz CERES 20040209/0007 130.8 11.03 OBJOd
1995, wasmadewith thisinstrument)butithasalso beenused Sky 19960629/0007 335.6 11.07 OBJO
for many other programs in stellar physics and galactic struc- Sky 20060613/0010 190.9 10.96 OBJOs
ture. The spectra cover the interval from 389.5 - 681.5 nm Sky 20060613/0011 286.3 10.98 OBJOs
and are recorded as 67 orders with a resolution of R≃42000 Sky 20021127/0009 430.1 10.95 OBJOd
(Baranneetal. 1996). The archive provides the spectra in two Sky 20060601/0027 142.4 11.04 OBJOd
formats. In the S2D format, the spectra are recorded as 67 × MOON 19980114/0008 381.4 11.06 OBJOd
1024pixelswiththecoefficientsofthepixeltowavelengthrela- MOON 19981001/0012 110.8 10.99 OBTHs
tionofeachorderbeingstoredintheFITSheader,anddeblazed MOON 19991222/0013 139.6 11.05 OBJOd
(theblazefunctionbeingstoredinaFITSextension).Inthespec MOON 19991222/0014 156.5 11.06 OBJOd
format,theordersarereconnectedandresampledinwavelength MOON 20000124/0028 200.0 11.05 OBJOd
with a constantstep of0.05Å, coveringthe range400-680nm MOON 20000124/0029 224.9 11.05 OBJOd
(thefirst3blueordersarenotincludedastheirsignaltonoisera- MOON 20000609/0008 350.8 11.08 OBJOd
tioisusuallyverylow).Thereareseveralmodesofobservation MOON 20000609/0009 246.4 11.08 OBJOd
relatedtothetwosetsofopticaltwin-fibers,directandscrambled VESTA 19950110/0020 98.7 11.11 OBTH
:onefiberisassignedtotheobjectwhilethesecondonecanbe CALLISTO 19990813/0037 218.2 11.10 OBJOd
masked,oronthesky,orilluminatedbyathorium-argonlamp.
TheELODIEarchivealsoreleasestheradialvelocitiescalculated
atthetelescopebycross-correlationofthespectrawithnumeri-
calmasksofspectraltypeF0orK0.Thegaussianfitofacross- 3. Minimumdistancebetweenspectra
correlationfunction(CCF)providesthefittedradialvelocity,the
To measure the degree of similarity between two spectra, of a
fullwidth at half maximum(FWHM) andthe amplitudeof the
target and of the Sun, we applied the TGMET code developed
correlationpeak.Inthecaseofspectroscopicbinaries,adouble
by Katzetal. (1998), a purely differential method implement-
gaussianisfitted.TheCCFsareprovidedasFITSfiles,withthe
ingaminimumdistancecriterion.TGMETwasalreadyapplied
parametersofthegaussianfitintheheader.Moredetailedinfor-
mation on the ELODIE archive data products is available in the bySoubiran&Triaud(2004)onaselectionof∼200ELODIEG
dwarfspectrainordertoidentifysolartwins.
on-lineuser’sguide1.
ThemethodwasappliedtoS2Dspectra,onorders21to67,
TheELODIEarchivecontainsmorethan35000publicspectra
andthereare∼8000distinctobjectnames(withpossibledupli- correspondingto the wavelengthrange 440–680Å, with rejec-
tionoftheunder-illuminatededgesoftheorders.Theorder63(∼
cation dueto unresolvedaliases). Forthis work onsolar twins,
onlyspectrawithasignaltonoiseratio(S/N)at550nmgreater 648–652.5nm)wasrejectedbecauseitisveryaffectedbytel-
luriclines.WebrieflyrecalltheprincipleoftheTGMETmethod.
than70,ameasuredradialvelocityandanidentificationresolved
Twospectratobecomparedneedtobeputonthesamescale,in
bySimbadwereselected.SpectrashowinganenlargedCCFwith
FWHMgreaterthan12kms−1oradouble-peakedCCFwerere- wavelengthandinflux.Thewavelengthalignementismadeby
shiftingthecomparedspectrumtotheradialvelocityofthesolar
jected. No otherselection based on colouror spectraltype was
spectrum and by resampling it to the same wavelength points.
1 http://atlas.obs-hp.fr/elodie/intro.html Thisoperationimpliesaninterpolationbetweenwavelengthsfor
Articlenumber,page2of15
D.Mahdietal.:SolartwinsintheELODIEarchive
the comparedspectrum, which is performedwith the quadratic (Blanco-Cuaresmaetal. 2014), following the recipes de-
Besselformula.Oncethehorizontalajustementisdone,thever- scribed in Blanco-Cuaresmaetal. (2015). For that task we
ticaloneisperfomedbyfittingthefluxofthecomparedspectrum used the ELODIE spectra in spec format (i.e. with the orders
tothesolarspectrum.Todoso,asimplefactorisdeterminedby reconnected and the wavelengths resampled at a constant
least-squares(thetwostarshavingroughlythesametemperature step) in the range 480-670 nm where iSpec was extensively
itisnotnecessarytointroduceaslope).Thereducedχ2 ofthat tested and validated. As model atmospheres we used MARCS
fit, computedorderbyorderovernearly40000wavelengthsin (Gustafssonetal.2008).
total, is thus a distance quantifying the similarity between the
Forthe atmosphericparameters,the line list establishedfor
two spectra. However following Katzetal. (1998), we did not
the Gaia ESO survey (Heiteretal. 2015) was used. It consists
adopt the real reduced χ2 for the distance between two spec-
of ∼ 140000 atomic lines extracted from the VALD database
tra, which would imply taking into account the noise on each
among which ∼2000 lines of 35 elements have been revisited
pixel.Instead,wecomputedanaveragedandnormalizedinstru-
andevaluatedonthebasisofthequalityoftheatomicdataand
mental response curve that we used as the weighting function
of the spectral synthesis for the Sun and Arcturus. For our use
of the fit. This smooth function reflects the global variation of
the recommended lines were selected, those flagged to be the
S/N over each order (the edges of an order are under-exposed
mostreliableones.T ,loggand[M/H]weredeterminedauto-
comparedtoitscentralpart).Itgivesasimilarweighttothecon- eff
maticallyanditeratively.Thedifferentstepsare:(1)thenoma-
tinuumandtothewingsandbottomofabsorptionlines,contrary
lizationofthespectrabytwodegreesplinesateverynanometer
toaweightingfunctionbasedonthephotonnoise.Severaltests
on pre-selected points, (2) a first guess of atmospheric param-
madeinKatzetal.(1998)demonstrateditshigherperformance,
eters by comparing the wings of the Hα and Hβ lines and the
especiallyathighS/N.
Mgtriplettoasmallpre-computedsyntheticgrid,(3)thegaus-
TGMET was applied to compute the distance of each so-
sianfitoflineswithrejectionofbadlyfittedlines(blendedlines,
lar spectrum to each of the other spectra. The TGMET output
gaps,cosmics,toofaintlines,..),(4)thesynthesisoftheremain-
foragivensolarspectrumresultedin2801distances(for2784
inglinesallowingthedeterminationoftheatmosphericparame-
target spectra plus the 17 other solar spectra) that were sorted
tersbyleastsquares.VmicandVmacweresetasfreeparameters
byincreasingvalue.Sincesolarspectrawereprocessedlikethe while vsini was set to 2 km s−1 for all the targets. These three
other targets, they were used to verify the results : the nearest
parametersarepossiblydegeneratedbuttheselectionofspectra
neighboursofagivensolarspectrumareexpectedtoalwaysbe withFWHM ≤ 12kms−1 preventedusfromdealingwithstars
theothersolarspectra.Thisisverifiedinmostcases,butsome-
rotatingmuchfasterthantheSunorwithhighermacroturbulent
timessomestarsarefoundclosertoagivensolarspectrumthan
velocity.Inpractice 615to 805lines of28 elementswere used
othersolarspectra,demonstratingthattheobservingconditions
forthe determinationof atmosphericparameters,dependingon
impacttheresults. Thisalsomeansthatsuchstarshavespectra
thespectrum.
very similar to that of the Sun, and thus they are good candi-
dates for solar twins. TGMET was also run with spectra con- In order to evaluate systematic errors and correctthem, we
volved to a common broadening of FWHM=12km s−1 in or- investigatedin detailthe resultsobtainedfor the 18 solar spec-
der to minimize the possible effect of a different rotation and tra.Averyhighconsistencywasobtained.ForTeff andlogg,the
macroturbulenceinthetargetsandintheSun.Thishoweverdid averagedvaluesare5773Kand4.32withstandarddeviationsof
not produce significantly different results. Other TGMET runs 9Kand0.02dexrespectively.TheTeff differencewiththefun-
wereperformedonasmallernumberoforders,thoseidentified damentalvalue,5777K,isnegligible.Itisworthtonotethatthe
tocarrymostinformation.Onerunwasperformedwithout10or- fundamentalTeffoftheSunwasrecentlyrevisedbyHeiteretal.
dersfoundtogivesystematicallylargerdistancesthantheother (2015)tobeTeff⊙=5771±1K,puttingourdeterminationeven
orders,andmorescatteredresults,amongthesolarspectra.An- closertothatfundamentalvalue.Wefoundasystematicshiftof
otherrunwasperformedusingonlytheorder39centeredonthe -0.12 dex for gravity with respect to the fundamental value of
Mg triplet at ∼ 520 nm, which is sensitive to effective temper- logg⊙ =4.44.Thisbiaswasfoundinpreviousworksbasedon
ature,surface gravityandmetallicity andthusa goodregionto iSpec(Blanco-Cuaresmaetal.2015).Acorrectionof+0.12dex
testthespectrocopicsimilarityofstars.Thebesttwincandidates was thus applied for the rest of the analyzed targets, establish-
were identified by mixing these results obtained with TGMET ing a global zero point centered on the solar reference param-
runsindifferentconfigurations.Intotal108TGMEToutputfiles eters. Similarly, we found an average metallicity [M/H]=-0.15
wereobtained,correspondingto18solarspectrax2(convolved dex for the solar spectra, determinedfrom lines of variousele-
and non convolvedspectra) x 3 (all orders, selected orders, or- ments. The global metallicities of the targets, determined from
der39).Thelistofsolartwin candidateswasbuiltbyselecting thesamelines,werethuscorrectedbythisvaluetoberelativeto
all stars ranked at a smaller distance than the last solar spec- theSun.
truminthesedifferentresultfiles.Thisgives56stars(225spec- AsstatedintheELODIEuser’smanual,itissuspectedthatfor
tra)amongwhichremarkablestarsareto benoted.HIP079672 image types OBTH, the presence of the thorium-argon orders
(HD146233,18Sco)isbyfarthemostfrequentfirstneighbour. interleaved with the stellar orders may lead to pollution of the
HIP018413(HD024409),HIP076114(HD138573),HIP089474 stellarspectrumbyhighlysaturatedArgonlineswhichcancon-
(HD168009) appear in the 5 nearest neighbours in more than taminatetheadjacentstellarorders.Soin principlethiskindof
70%oftheresultfiles,followedbyHIP118162(HD224465)and spectrashouldnotbeusedforthedeterminationofatmospheric
HIP041484(HD071148).Thestatusofthesestarsassolartwins parametersandabundances.Howeverweobtainedsimilaratmo-
willbediscussedinSect.8. sphericparametersfortheSunwhenweaveragedthemfromthe
15OBJand3OBTHspectraseparately:negligeabledifferences
of 8K, 0.02 dex and 0.01 dex were measured in T , logg and
4. Atmosphericparameters eff
[M/H] respectively. There is thus apparently no impact of the
The atmospheric parameters and abundances of the imagetypeontheatmosphericparameters,probablyduetothe
solar and target spectra were derived with iSpec selectionofwellfittedlinesandtothesynthesismethod.
Articlenumber,page3of15
A&Aproofs:manuscriptno.solar_twins_AA
In the target sample, 48 stars have dupplicate observations
fromwhichweestimatedtheinternalconsistencyofthederived
atmospheric parameters. For those stars the standard deviation
around the mean Teff ranges from 1 K to 20K, with a median
of9K.Forloggthemedianstandarddeviationis0.02dex,with
nohighervaluethan0.04.For[M/H]allthestarshaveamedian
standard deviationof 0.01dex, exceptHIP097336which has a
standarddeviationof0.02dexfor5spectra.Althoughtheseval-
ues show the excellentconsistencyof the iSpec determinations
ofatmosphericparametersfromonespectrumtoanother,westill
verified that the 16 stars observed with both modes OBJ and
OBTH have derived parameters in good agreement from both
typesofexposures.
HIP079672 (18 Sco) is one of the Gaia benchmark stars
recently studied by Heiteretal. (2015) who determined fun-
damental T =5810 K and logg=4.44 from the defining rela-
eff
tions,independentlyofspectroscopy.TheGaiabenchmarkstars
are intended to serve as calibrators for spectroscopy and thus
can beused to evaluateourdeterminations.Only onespectrum
is available for HIP079672 in the ELODIE archive from which
wedeterminedT =5793±9Kandlogg=4.43±0.015,ingood
eff
agreement with the fundamental values. We thus confirm that
HIP079672isslightlyhotterthantheSun.
Atmospheric parameters are presented in Table 2. The his-
tograms of the differences with the solar values are presented
inFig.1.Thereare22starswhichfallintothecategoryofsolar
twinaccordingtothecriteriaadoptedbyMeléndezetal.(2014b)
: they differ by less than 100 K in T , 0.1 dex in logg, and
eff
0.1 dex in metallicity from the Sun values. We flag such stars
with AP category= B in Table 2. Three remarkablestars have
tobenotedastheyhavethesameatmosphericparametersasthe
Sun,within25KinT and0.05dexinloggand[M/H](flagged
eff
withAPcategory=AinTable2):HIP076114,HIP085244and
HIP088194.Thestatusofthesestarsassolartwinswillbedis-
cussedinSect.8.
Interestinglysomestarshavequitedifferentparametersfrom
theSun(flaggedwithAPcategory=CinTable2).Forinstance
HIP065721isasub-giantcoolerthantheSunandslightlymore
metal-poor. This example illustrates the degeneracy of atmo-
sphericparametersatthe resolutionof42000whenthe overall
spectrumisconsidered.Thissuggeststhatthespectrocopiccom-
parisonbetweenspectrashouldbeperformeddifferentially,ona
linebylinebasis.
5. Licontent
Itisknownthatthelithiumabundancesofsolartypestarsshow Fig.1.Histogramsofthedifferencesbetweentheatmosphericparame-
alargedispersion(seeforinstancethecomprehensivestudyby tersofthetwincandidatesandoftheSun,asobtainedbyiSpec.
Takedaetal. 2007). This is subject to various interpretations.
The Sun is depleted in Li and it is expected that solar twins
show the same deficiency. We have classified our targets stars
starsexaminedherearesupposedlyslowrotatorsduetoourini-
intothreecategoriesafteravisualinspectionofthewavelength
rangeinludingtheLilinesat∼670.78nmincomparisontoone tialconstraintFWHM≤12kms−1.Itisthusinterestingtofind
sodiversestrengthsoftheirLifeature.
ofthe ELODIEsolarspectra.We classified into A inTable 2 the
starswhichshowasimilarLideficiencyastheSun,intoBthose
withaslightlyhigherLiabundanceandintoCthosewhichex-
6. Abundances
hibit a pronouncedLi feature. There are 24, 11 and 21 stars in
categories A, B and C respectively.Figure 2 shows two exam- Thechemicalabundanceswerederivedwith iSpecandthe line
plesofstarsineachcategory.ItisworthtonotethatHIP076114 list of the Gaia ESO survey, this time including also lines la-
andHIP088194areclassifiedAfortheirsimilaritytotheSunin beledtobelessreliable,duetouncertainatomicdataorpossible
bothatmosphericparametersandLithiumcontent. blends.Howeverthe18solarspectraallowedustokeeponlythe
Stellar rotationandactivity are stronglycorrelatedwith the lines which gave consistent abundances from one spectrum to
surface Li content of solar type stars (Takedaetal. 2007). The another.Moreoversince we are dealingwith solar twins, a dif-
Articlenumber,page4of15
D.Mahdietal.:SolartwinsintheELODIEarchive
Fig.2.Comparisonofthespectralrangecenteredonthe670.78nmLifeatureinoneoftheSunspectra(blackthickline)andatarget(greendashed
line),withupto5spectraavailableforsomestars.TheupperpanelsshowtwostarsclassifiedA(sameLideficiencyastheSun),themiddlepanels
showtwostarsclassifiedB(slightlymoreabundantinLithantheSun),thebottompanelsshowtwostarsclassifiedC(pronouncedLifeature).
ferentialanalysiswith respecttosolar spectrawaspossible,re- theELODIEarchiveinfavourofstarsfollowed-upforthesearch
ducingtheuncertaintyinrelativeabundancesduetobadatomic ofextra-solarplanets.
data.Inthatwaywewereabletoobtainahighprecision,which
issoimportanttoidentifysolartwins(Meléndezetal.2012).
Atotalof1865linesweremeasurableintheELODIEspectra
outofwhichwerekeptonlythosemeasuredinatleast17outof
18 solar spectra with a weighted standard deviation lower than
0.02.Thenweselectedthechemicalelementsforwhichatleast
3 such lines were measured. This left us with 189 Fe lines, 23
Nilines,15SiandCalines,8Crlines,5MnandTilines,4Na
lines, all neutral,plus3 FeII lines. Table 7 lists these 267lines
which were found to be suitable for spectral synthesis of solar
type stars. For each of these lines, we took as reference abun-
dancetheweightedmeanobtainedonthe17or18solarspectra.
The lineby line abundancesofthe 225spectrawere thenmea-
sureddifferentiallyforthisset.Asthetotalerrorforalineabun-
dancerelativetotheSun,wequadraticallysummedtheweighted
standarddeviationobtainedforthesolarspectraandthermsof
thefitobtainedforthetargetspectrum.Thentheweightedaver-
ageandstandarddeviationwerecomputedforeachelementand Fig.3.Differenceofironabundancesobtainedwithneutralandionised
eachstar,aspresentedinTables4,5and6.Withupto5spectra linesforthe56twincandidates.Themeandifferenceis0.008dexwith
astandarddeviationof0.019dex.
availableforalargefractionofthetargets,thedeterminationof
[FeI/H] was often based on the measurementof 945individual
lines.Intotalweobtained59707individuallinemeasurements
HIP079672 (18 Sco) has been studied at very high resolu-
witherrorsbetween0.01and0.111andamedianerrorof0.02.
tion (R ∼ 110000) and S/N (800 - 1000) by Meléndezetal.
Asaverification,we checkedtheagreementofthe[FeI/H]and
(2014a)whoobtainedabundancesdifferentiallytotheSunwith
[FeII/H] determinations(Fig. 3). The meandifferenceis 0.008,
anunprecedentedlevelofprecision.Wecomparetheirdetermi-
demonstrating the excellent agreement of the 2 determinations
nations to ours in Table 3 : they agree within 0.03 for Ti and
ofironabundances,withastandarddeviationof0.019,agreeing
Cr,within0.008forNaandCaandwithin0.004forFe,Si,Mn
perfectlywiththemedianerrorof0.02ofalltheindividualline
andNi.Thustheagreementislargelybetterthanexpectedfrom
measurements.
ourquotederrors.Forcomparisonwealsoshowtheabundances
Fig.3alsoshowsthatthereisaconcentrationofstarsslightly fromJofréetal.(2015)whomadeanextensivestudyoftheGaia
more metal-rich than the Sun, a possible bias of the contentof benchmarkstars.Theyarealsoinexcellentagreement.
Articlenumber,page5of15
A&Aproofs:manuscriptno.solar_twins_AA
Table 3. Comparison of abundances determined in this work for HIP089474, HIP118162, HIP041484. Three stars were found
HIP079672(18Sco)tothoseobtainedbyMeléndezetal.(2014a)and
tohavesolaratmosphericparameters:HIP076114,HIP085244
byJofréetal.(2015).
andHIP088194,twoofwhichwithalsothesameLideficiency
: HIP076114, HIP088194. Seven stars were found to have the
[X/H] Thiswork Meléndezetal.(2014a) Jofréetal.(2015)
solar abundance pattern for 8 chemical elements : HIP021436,
Fe 0.051±0.026 0.054±0.005 0.03±0.01(a)
HIP035265,HIP076114,HIP085244,HIP089474,HIP100963,
Na 0.017±0.045 0.025±0.004
TYC2694-00364-1.Table9summarizesthecategoriesA,Band
Si 0.043±0.014 0.047±0.003 0.048±0.018
C that we gave to each target according to its similarity to the
Ca 0.065±0.016 0.057±0.006 0.058±0.036
Sun in atmospheric parameters, abundances, Li content, abso-
Ti 0.022±0.045 0.051±0.007 0.046±0.026
lutemagnitudeandage.
Cr 0.034±0.026 0.056±0.006 0.049±0.023
Itis remarkableto finda star similar to the Sunin allthese
Mn 0.037±0.012 0.041±0.006 0.040±0.018
properties:HIP076114(HD138573),supportedalsobytheab-
Ni 0.038±0.027 0.041±0.004 0.039±0.017
solute magnitude and most probable age range. This star is
also in the list of best twins by Datsonetal. (2012) although
Notes.(a)NLTEdeterminationfromJofréetal.(2014)
theyclassifiyitafterHIP079672(18Sco)(seealsoDatsonetal.
2015). HIP076114is also partof the surveyof solar twin stars
Seven solar analogs have all the tested elements agreeing within 50 parsecs of the Sun by PortodeMelloetal. (2014),
withthesolarabundanceswithin0.05:HIP021436,HIP035265, as well as in the Solar Twin Planet Search by Ramírezetal.
HIP076114, HIP085244, HIP089474, HIP100963, TYC2694- (2014b), but not mentioned as a remarkable twin in these two
00364-1. papers. It is worth noting that the galactic velocity of that star
Figure 4 shows the iron abundancesof the 56 targets from differs by ∼40 km s−1 from that of the Sun, suggesting that it
neutralandionizedlinesandFig. 5 displaysthe abundancesof is not a solar sibling. It is however an excellent star to use for
theotherelements.Thestarsarenumberedfrom1to56forthe calibrationsortosearchforexoplanets.
sakeofclarity.Theabundancesofthetestedelementsarequite
HIP085244isoursecondbesttwincandidate,withsolarat-
wellcenteredonthesolarvalue,withasmallspread.Thissug-
mosphericparametersandchemicalpatternbutitisclassifiedB
geststhatthesolarabundancepatternisnotunusualassuggested
fortheLi,withaslightlymoreenhancedcontent.Accordingto
bysomestudies(Nissen2015).Thispointneedshoweverfurther
XHIPthisstarissupposedtobeolderthantheSun.HIP085244
confirmationwithotherelementswhicharenotpartofthisstudy
isalsopartofthesurveyofsolartwinstarswithin50parsecsof
becauseofourstrictselectionoflinesverywellmeasuredinthe
theSunbyPortodeMelloetal.(2014).
ELODIEsolarspectra,whichhasguaranteeda highlevelofpre-
HIP088194hassolaratmosphericparametersandisdeficient
cision.
in Li, but it is slightly more metal-poor than the Sun. This is
clearly visible in the Figures 4 and 5 (star 40). It is older than
7. Additionalinformation the Sun according to XHIP, but of the same age according to
PortodeMelloetal.(2014).Itisoneofthenewcandidatescon-
The 56 twin candidates are either Hipparcos or Tycho stars.
sideredasexcellentinthatwork.
WehaveretrievedinformationsforthemintheXHIPcatalogue
HIP021436hasthesolarchemicalpattern,andalsothesame
(Anderson&Francis2012)andinSimbad,reportedinTable8.
Licontentbutits effective temperatureis lower by∼60 K than
All the stars are within 60 pc from the Sun, their spectral type
thatof the Sun. Despite us classifying it B for the atmospheric
varyfromF8toG6.5.TheirB-VcoloursandM absolutemag-
v
parameters, it still falls into the solar twin category according
nitudesspanawiderangeofvalues.Agesareprovidedinform
tothecriteriondefinedbyMeléndezetal.(2014b).OurT de-
ofaprobablerange,sometimesverylargeduetothelackofcon- eff
termination(5715±10K) is in verygoodagreementwith the 7
straintonthisfundamentalproperty.
valueslistedinthePASTELcatalog(Soubiranetal.2010)rang-
Weproceededlikeintheprevioussectionsandclassifiedthe
starsintotwocategoriesdependingonthesimilarityoftheirM ing from 5675 to 5748 K. It is worth noting that this star has
v
a galactic velocity compatible with that of the Sun within 10
absolute magnitudeand age with the solar ones. Stars with M
intherange[4.6–5.0]arepotentiallysimilartotheSun(M ≃v kms−1.ItsagereportedinXHIPisnotwellconstrained,sopos-
4.80).IftheirageintervalalsoincludestheageoftheSun(∼v⊙4.5 siblycompatiblewiththatoftheSun.Itisthusalsoaverygood
solarsiblingcandidate.
Gyr),thensuchstarsaregoodtwincandidates.Fifteenstarswere
foundtobesimilartotheSun,basedontheirabsolutemagnitude HIP035265hasthesolarchemicalpattern,anabsolutemag-
andagerange. nitudeandagecompatiblewiththe Sun,butitis hotterby∼80
We have also searched in Simbad whether the twin candi- KandwithastrongLifeature.Italsohasagalacticvelocitydif-
dateshavea detectedextrasolarplanetandfound4such stars: ferentfromthe Sun.Thisis thusneithera perfectsolar twinor
HIP053721,HIP065721,HIP096901,HIP097336,noneofthem sibling.
beingagoodsolartwin. HIP100963 has the solar chemical pattern, but it is hotter
than the Sun (T = 5821±6 K) and has a pronounced Li fea-
eff
ture.ItwaspreviouslyidentifiedassolartwinbyRamírezetal.
8. Discussion
(2009)whodeterminedT =5815Kand[Fe/H]=+0.018±0.019
eff
In each of the previous sections, we have classified 56 twin inverygoodagreementwithourdeterminations(wedetermined
candidatesaccordingto their similarity with the Sun using dif- [Fe/H]=+0.011±0.034).Consideringthisstarasanexcellentso-
ferent properties : global spectrum, atmospheric parameters, lar twin, Takeda&Tajitsu (2009) made an extensive study at
Li content, abundances of 8 chemical elements, absolute mag- very high resolution (R=90000) and high S/N (500-1000)and
nitude, age, presence of a planet. According to the TGMET foundit hotter than the Sun by 23 K, with a difference of iron
method, 6 stars were found to have a spectrum very simi- abundanceof0.004andahigherLicontentbyafactorof∼56as
lar to that of the Sun : HIP079672, HIP018413, HIP076114, comparedtoSun,alsoingoodagreementwithourfindings.
Articlenumber,page6of15
D.Mahdietal.:SolartwinsintheELODIEarchive
TYC2694-00364-1hasbeenpoorlystudiedasanindividual ELODIEarchiveisaverygoodresourcetomakecomprehensive
star until now. It essentially differs from the Sun by its hotter studiesofsolartypestars.
temperature(T =5842±1K)andastrongLifeature. WefoundthatthebestsolartwininoursampleisHIP076114
eff
HIP042575,HIP056832and HIP118162 are our next good (HD138573), dethroning previous candidates (HIP079672/18
solar twins, never identified as such before. Although slightly Sco,HIP100963).AlltheotherstarsslightlydifferfromtheSun
metal-rich, their abundances differ by less than 0.1 dex from in one or another property. We list 19 solar twins which have
those of the Sun. They exhibit the solar Li deficiency and Teff, logg and abundances differing from those of the Sun by
have an age and absolute magnitude compatible with those of lessthan100Kand0.1dexrespectively.
the Sun. Next is HIP011728, also more metal-rich but previ- Wefoundagoodsolarsiblingcandidate:HIP021436which
ously identified as solar twin by PortodeMelloetal. (2014) wouldbeworthstudyingfurther.
and Datsonetal. (2015). Its absolute magnitude is reported as
Acknowledgements. CSwouldliketodedicatethispapertothememoryofGiusa
Mv=5.03 in XHIP, higher than that of the Sun. Then we have CayreldeStrobel,aninspirationforayoungergenerationofastronomersinthe
HIP018413,a new solar twin, slightly colder,more metal-poor questforsolartwinsandmanyothertopics.DMacknowledgesfinancialsupport
and older than the Sun, which was one of the closest twins ac- from the Irakian ministry ofhigher education and research. Weacknowledge
constructivecommentsandsuggestionsbytheanonymousreferee.Weusedthe
cordingto TGMET.HIP049756was alreadyidentifiedas a so-
CDS-SIMBADandNASA-ADSdatabases,andtheVizieRserviceatCDS.
lar twin in several previous studies. We classified it B for the
atmospheric parametersand abundancesbecause of [M/H] and
[Ca/H]beingslightlyhigherthanourlimits,0.05dexand0.058
References
dex respectively, while all the other values are solar. Its main
differencetotheSunisahigherlithiumcontent. Anderson,E.,&Francis,C.2012,AstronomyLetters,38,331
Baranne,A.,Queloz,D.,Mayor,M.,etal.1996,A&AS,119,373
The most studied HIP079672 (18 Sco) is definitively not a
Batista,S.F.A.,Adibekyan,V.Z.,Sousa,S.G.,etal.2014,A&A,564,A43
perfectsolartwin,accordingtoourrankingatthe14thposition Bazot,M.,Ireland,M.J.,Huber,D.,etal.2011,A&A,526,L4
Blanco-Cuaresma, S., Soubiran, C., Heiter, U., & Jofré, P. 2014, A&A, 569,
in Table 9. We find a hotter temperature and enhanced abun-
A111
dancesrelativetotheSun,includingthelithium,thusconfirming Blanco-Cuaresma,S.,Soubiran,C.,Heiter,U.,etal.2015,A&A,577,A47
thefindingsofMeléndezetal.(2014a). Casagrande, L.,Ramírez, I.,Meléndez, J.,Bessell, M.,&Asplund, M.2010,
A&A,512,
Thenextfoursolar twinsare alreadyknown.Theyhaveall Casagrande,L.,Portinari,L.,Glass,I.S.,etal.2014,MNRAS,439,2060
ahigherlithiumcontent.Thelastoneofthelist,HIP094981,is CayreldeStrobel,G.1996,A&ARev.,7,243
CayreldeStrobel,G.,&Bentolila,C.1989,A&A,211,324
unknown.Itis colderand moremetalrichbutstill a solartwin CayreldeStrobel,G.,Knowles,N.,Hernandez,G.,&Bentolila,C.1981,A&A,
according to the criterion of Meléndezetal. (2014b). It is also 94,1
Datson,J.,Flynn,C.,&Portinari,L.2012,MNRAS,426,484
more luminous (Mv=4.27) and much older than the Sun (10.4 Datson,J.,Flynn,C.,&Portinari,L.2014,MNRAS,439,1028
Gyr)withastrongLifeature. Datson,J.,Flynn,C.,&Portinari,L.2015,A&A,574,A124
denHartog,E.A.,Ruffoni,M.P.,Lawler,J.E.,etal.2014,ApJS,215,23
The second part of Table 9 lists the other stars which were Friel,E.,CayreldeStrobel,G.,Chmielewski,Y.,etal.1993,A&A,274,825
found to differ from the Sun, either in their atmospheric pa- Gray,R.O.,&Corbally,C.J.1994,AJ,107,742
Grevesse,N.,Asplund,M.,&Sauval,A.J.2007,SpaceSci.Rev.,130,105
rameters or in their abundances. It is however worth to note Gustafsson,B.,Edvardsson,B.,Eriksson,K.,etal.2008,A&A,486,951
Hardorp,J.1978,A&A,63,383
HIP007339andHIP089474whichhavethesolarchemicalpat-
Heiter,U.,Lind,K.,Asplund,M.,etal.2015,Phys.Scr,90,054010
ternandlithiumdeficiency.HIP007339isoneofthecoldeststar Heiter,U.,Jofré,P.,Gustafsson,B.,etal.2015,arXiv:1506.06095
Jofré,P.,Heiter,U.,Soubiran,C.,etal.2014,A&A,564,A133
ofoursample(T =5630±12K)whileHIP089474differsfrom
eff Jofré,P.,Heiter,U.,Soubiran,C.,etal.2015,arXiv:1507.00027
the Sun by its lower surface gravity (logg= 4.29±0.01). Both Johansson,S.,Litzén,U.,Lundberg,H.,&Zhang,Z.2003,ApJ,584,L107
Katz,D.,Soubiran,C.,Cayrel,R.,Adda,M.,&Cautain,R.1998,A&A,338,
stars are not however good solar sibling candidates, having a
151
galacticvelocitysignificantlydifferentfromtheSun,witha ra- Kurucz,R.L.2008,RobertL.Kuruczon-linedatabaseofobservedandpredicted
dialvelocitycomponentU=50.8kms−1forHIP007339andaro- atomictransitions
Li,T.D.,Bi,S.L.,Liu,K.,Tian,Z.J.,&Shuai,G.Z.2012,A&A,546,A83
tationalcomponentV=-62.1kms−1forHIP089474.HIP089474 Liu,C.,Ruchti,G.,Feltzing,S.,etal.2015,A&A,575,A51
is also old (9.1 Gyr), which interestingly makes it a possible May,M.,Richter,J.,&Wichelmann,J.1974,A&AS,18,405
Mayor,M.,&Queloz,D.1995,Nature,378,355
memberofthethickdisk. Meléndez,J.,Dodds-Eden,K.,&Robles,J.A.2006,ApJ,641,L133
Meléndez,J.,&Ramírez,I.2007,ApJ,669,L89
Meléndez,J.,Bergemann,M.,Cohen,J.G.,etal.2012,A&A,543,A29
Meléndez,J.,Ramírez,I.,Karakas,A.I.,etal.2014,ApJ,791,14
Meléndez,J.,Schirbel,L.,Monroe,T.R.,etal.2014,A&A,567,L3
9. Conclusion
Monroe,T.R.,Meléndez,J.,Ramírez,I.,etal.2013,ApJ,774,L32
Moultaka,J.,Ilovaisky,S.A.,Prugniel,P.,&Soubiran,C.2004,PASP,116,693
Inthiswork,wehaveselected56solartwincandidates,among Nakajima,T.,&Morino,J.-I.2012,AJ,143,2
Nissen,P.E.2015,A&A,579,A52
1165testedstars,basedonthesimilarityoftheirspectrumtothat O’Brian, T.R.,Wickliffe, M.E.,Lawler, J.E.,Whaling, W.,& Brault, J.W.
of the Sun. We have determined their atmospheric parameters 1991,JournaloftheOpticalSocietyofAmericaBOpticalPhysics,8,1185
Önehag,A.,Korn,A.,Gustafsson,B.,Stempels,E.,&Vandenberg,D.A.2011,
andabundances,andexaminedtheirLicontenttofurtherstudy
A&A,528,A85
theirsimilaritytotheSun. Pichardo,B.,Moreno,E.,Allen,C.,etal.2012,AJ,143,73
PortodeMello,G.F.,&daSilva,L.1997,ApJ,482,L89
A strength of our study is that we have used as reference
PortodeMello,G.F.,daSilva,R.,daSilva,L.,&deNader,R.V.2014,A&A,
a large number of solar spectra (18), representive of the range 563,A52
of observing conditions, and from which we have selected the Puspitarini,L.,Lallement,R.,Vergely,J.-L.,&Snowden,S.L.2014,A&A,566,
A13
best lines to be analyzed for abundancemeasurements by syn-
Ramírez,I.,Meléndez,J.,&Asplund,M.2009,A&A,508,L17
thesis.Severalspectrawereavailableforthemajorityofthese- Ramírez,I.,Bajkova,A.T.,Bobylev,V.V.,etal.2014,ApJ,787,154
Ramírez,I.,Meléndez,J.,Bean,J.,etal.2014,A&A,572,A48
lected targets, up to 5 (225 spectra in total). The analysis was Ruffoni,M.P.,denHartog,E.A.,Lawler,J.E.,etal.2014,MNRAS,441,3127
performeddifferentiallytothesolarspectra,onalinebylineba- Soubiran,C.,&Triaud,A.2004,A&A,418,1089
Soubiran,C.,LeCampion,J.-F.,CayreldeStrobel,G.,&Caillo,A.2010,A&A,
sisfortheselectedsetoflines.Thisledtoaverygoodinternal
515,A111
precision. Our atmospheric parameters and abundances proved Takeda,Y.,Kawanomoto,S.,Honda,S.,Ando,H.,&Sakurai,T.2007,A&A,
to also be accurate, in excellent agreement with other studies 468,663
Takeda,Y.,&Tajitsu,A.2009,PASJ,61,471
made at higher resolution and S/N. This demonstrates that the
Articlenumber,page7of15
A&Aproofs:manuscriptno.solar_twins_AA
Table2.AtmosphericparametersobtainedwithiSpec.loggand[M/H]werescaledtothesolarvalues.Inthefifthcolumn,Nisthenumberof
spectraavailableforthestar,fromwhichtheaveragedatmosphericparameterswerecomputed.Forthestarswithmultiplespectra,theerrorof
eachparameterisitsstandarddeviationaroundthemean.Forthe9starswithonesingleobservation,theerroristhatprovidedbyiSpec.Thelast
twocolumnsindicatewhetherthestarlooksverysimilar(A),similar(B)ordifferent(C)totheSuninitsatmosphericparameters(AP)andLi
featuresat∼670.78nm.
HIP T logg [M/H] N AP Li
eff
HIP001813 5772±19 4.30±0.03 -0.03±0.01 5 C A
HIP004290 5801±9 4.58±0.02 -0.07±0.01 5 C C
HIP007339 5630±12 4.36±0.02 0.02±0.01 1 C A
HIP007585 5779±10 4.41±0.01 0.07±0.01 5 B C
HIP007902 5631±11 4.36±0.02 -0.02±0.01 1 C A
HIP007918 5889±4 4.33±0.01 0.05±0.01 5 C C
HIP011253 5744±16 4.35±0.02 0.01±0.01 5 B B
HIP011728 5763±2 4.47±0.01 0.06±0.01 2 B A
HIP018413 5738±16 4.47±0.02 -0.04±0.01 5 B A
HIP021010 5533±14 3.85±0.02 -0.12±0.01 5 C C
HIP021436 5715±10 4.46±0.02 0.01±0.01 5 B A
HIP024813 5900±12 4.31±0.02 0.13±0.01 1 C C
HIP029432 5766±13 4.44±0.02 -0.09±0.01 5 B B
HIP029525 5751±8 4.60±0.01 -0.02±0.01 5 C C
HIP031965 5782±15 4.28±0.02 0.05±0.01 5 C A
HIP035265 5857±3 4.47±0.01 0.03±0.01 5 B C
HIP036874 5743±18 4.21±0.03 -0.07±0.01 2 C A
HIP041484 5855±11 4.40±0.02 0.03±0.01 5 B C
HIP041844 5860±19 4.25±0.02 -0.03±0.01 5 C B
HIP042575 5710±11 4.50±0.02 0.05±0.01 1 B A
HIP043557 5848±13 4.48±0.02 -0.04±0.01 1 B B
HIP043726 5776±10 4.54±0.02 0.13±0.01 1 C C
HIP044089 5749±9 4.31±0.02 0.04±0.01 5 C A
HIP049756 5787±14 4.41±0.02 0.05±0.01 5 B B
HIP050316 5779±9 4.07±0.02 0.02±0.01 5 C C
HIP050505 5661±5 4.49±0.01 -0.14±0.01 5 C A
HIP053721 5861±13 4.26±0.01 0.01±0.01 5 C C
HIP056832 5750±13 4.51±0.02 0.06±0.01 5 B A
HIP062527 5858±7 4.33±0.01 0.16±0.01 5 C B
HIP063636 5783±9 4.58±0.01 -0.04±0.01 5 C C
HIP064150 5726±20 4.32±0.03 0.06±0.01 5 C A
HIP065721 5501±5 3.90±0.01 -0.09±0.01 5 C C
HIP072043 5811±18 4.27±0.02 -0.02±0.01 5 C A
HIP076114 5757±10 4.42±0.01 0.00±0.01 5 A A
HIP077052 5673±13 4.52±0.01 0.07±0.01 5 C B
HIP079672 5797±9 4.43±0.02 0.05±0.01 1 B B
HIP085244 5789±13 4.47±0.02 0.04±0.01 1 A B
HIP085810 5862±14 4.45±0.02 0.17±0.01 5 C C
HIP086193 5677±9 4.26±0.02 0.06±0.01 5 C A
HIP088194 5759±4 4.40±0.01 -0.05±0.01 5 A A
HIP089474 5802±10 4.29±0.01 0.02±0.01 5 C A
HIP090355 5585±3 4.45±0.01 -0.14±0.01 3 C A
HIP094981 5693±7 4.52±0.01 0.06±0.01 5 B C
HIP096402 5612±2 4.17±0.01 -0.06±0.01 2 C A
HIP096895 5827±12 4.36±0.02 0.15±0.01 1 C A
HIP096901 5735±9 4.25±0.04 0.06±0.01 2 C B
HIP096948 5761±14 4.39±0.01 0.11±0.01 3 C B
HIP097336 5812±13 4.36±0.01 0.13±0.02 5 C A
HIP097420 5798±11 4.50±0.02 0.03±0.01 5 B C
HIP097767 5670±5 4.03±0.01 -0.16±0.01 5 C A
HIP100963 5821±6 4.49±0.01 0.01±0.01 5 B C
HIP102040 5840±4 4.45±0.01 -0.07±0.01 5 B C
HIP116613 5832±10 4.57±0.01 0.13±0.01 5 C C
HIP118162 5753±4 4.44±0.01 0.08±0.01 2 B A
TYC2583-01846-2 5890±6 4.53±0.01 -0.03±0.01 5 C C
TYC2694-00364-1 5842±1 4.44±0.01 0.03±0.01 5 B C
Articlenumber,page8of15
D.Mahdietal.:SolartwinsintheELODIEarchive
Fig.4.Ironabundanceofthe56solartwincandidates,obtainedfromneutralandionizedlines.Reddotsrepresentthe7starsfoundtohavethe
sameabundancesthantheSun,within0.05dex,forallthetestedelements.
Fig.5.LikeFig.4fortheotherelements.
Articlenumber,page9of15
A&Aproofs:manuscriptno.solar_twins_AA
Table4.Detailedabundancesofthe56targets
num HIP/TYC [FeI/H] [FeII/H] [Na/H]
Mean σ N Mean σ N Mean σ N
1 HIP001813 -0.024 0.038 928 -0.036 0.029 15 -0.076 0.052 16
2 HIP004290 -0.062 0.032 938 -0.055 0.036 15 -0.171 0.067 17
3 HIP007339 0.016 0.038 185 -0.058 0.132 3 0.044 0.028 3
4 HIP007585 0.069 0.039 926 0.067 0.043 15 0.032 0.044 15
5 HIP007902 -0.021 0.036 187 -0.080 0.105 3 -0.112 0.041 4
6 HIP007918 0.049 0.028 944 0.043 0.024 15 0.055 0.031 20
7 HIP011253 0.001 0.044 940 0.004 0.028 15 0.094 0.054 19
8 HIP011728 0.049 0.033 376 0.039 0.043 6 0.045 0.026 8
9 HIP018413 -0.043 0.035 939 -0.033 0.046 15 -0.015 0.044 18
10 HIP021010 -0.116 0.037 940 -0.134 0.041 15 -0.126 0.042 20
11 HIP021436 0.011 0.039 939 -0.017 0.053 15 -0.006 0.047 18
12 HIP024813 0.121 0.028 187 0.124 0.018 3 0.126 0.080 4
13 HIP029432 -0.080 0.033 933 -0.080 0.047 15 -0.125 0.065 16
14 HIP029525 -0.021 0.035 939 -0.012 0.035 15 -0.109 0.040 19
15 HIP031965 0.047 0.035 935 0.031 0.059 15 0.056 0.043 20
16 HIP035265 0.031 0.033 934 0.021 0.065 15 -0.032 0.068 20
17 HIP036874 -0.068 0.035 366 -0.096 0.166 6 -0.124 0.039 8
18 HIP041484 0.037 0.028 940 0.033 0.034 15 -0.011 0.044 18
19 HIP041844 -0.021 0.039 938 -0.048 0.079 15 0.002 0.097 20
20 HIP042575 0.050 0.031 186 0.037 0.013 3 0.003 0.078 4
21 HIP043557 -0.047 0.033 189 -0.044 0.023 3 -0.059 0.044 4
22 HIP043726 0.119 0.028 188 0.108 0.008 3 0.139 0.040 4
23 HIP044089 0.041 0.032 943 0.049 0.040 15 0.090 0.039 19
24 HIP049756 0.042 0.037 938 0.020 0.034 15 0.034 0.047 18
25 HIP050316 0.024 0.034 934 0.032 0.041 15 0.038 0.045 20
26 HIP050505 -0.137 0.029 945 -0.146 0.033 15 -0.174 0.039 21
27 HIP053721 0.018 0.030 940 0.022 0.030 15 0.054 0.037 20
28 HIP056832 0.054 0.040 938 0.063 0.049 15 0.005 0.041 20
29 HIP062527 0.147 0.033 940 0.132 0.033 15 0.195 0.038 19
30 HIP063636 -0.033 0.037 941 -0.033 0.046 15 -0.149 0.076 20
31 HIP064150 0.053 0.038 939 0.021 0.047 15 0.076 0.050 20
32 HIP065721 -0.091 0.033 939 -0.088 0.023 15 -0.129 0.044 19
33 HIP072043 -0.023 0.035 942 -0.043 0.039 15 -0.004 0.047 19
34 HIP076114 0.001 0.035 928 -0.015 0.030 15 -0.005 0.063 18
35 HIP077052 0.064 0.033 942 0.079 0.030 15 -0.018 0.046 20
36 HIP079672 0.051 0.026 188 0.051 0.032 3 0.017 0.045 4
37 HIP085244 0.035 0.034 188 0.048 0.010 3 0.021 0.031 4
38 HIP085810 0.166 0.033 923 0.157 0.039 15 0.160 0.027 18
39 HIP086193 0.052 0.040 923 0.055 0.052 15 0.030 0.045 18
40 HIP088194 -0.050 0.034 945 -0.085 0.048 15 -0.070 0.040 20
41 HIP089474 0.018 0.031 925 0.016 0.022 15 0.020 0.029 18
42 HIP090355 -0.152 0.033 566 -0.130 0.048 9 -0.158 0.071 12
43 HIP094981 0.060 0.032 945 0.068 0.028 15 -0.004 0.032 20
44 HIP096402 -0.081 0.036 377 -0.110 0.029 6 -0.022 0.042 8
45 HIP096895 0.133 0.026 184 0.108 0.027 3 0.149 0.016 3
46 HIP096901 0.062 0.036 374 0.047 0.030 6 0.073 0.031 6
47 HIP096948 0.100 0.044 555 0.070 0.038 9 0.073 0.057 11
48 HIP097336 0.119 0.037 941 0.108 0.037 15 0.144 0.041 18
49 HIP097420 0.037 0.033 936 0.065 0.033 15 -0.070 0.036 17
50 HIP097767 -0.164 0.033 940 -0.179 0.036 15 -0.156 0.055 19
51 HIP100963 0.011 0.034 946 -0.015 0.031 15 -0.031 0.034 20
52 HIP102040 -0.070 0.030 932 -0.079 0.038 15 -0.161 0.022 17
53 HIP116613 0.124 0.034 944 0.157 0.024 15 0.011 0.030 19
54 HIP118162 0.073 0.029 378 0.072 0.021 6 0.077 0.113 8
55 TYC2583-01846-2 -0.027 0.028 943 -0.030 0.056 15 -0.135 0.058 19
56 TYC2694-00364-1 0.027 0.033 933 0.028 0.041 15 -0.040 0.022 15
Articlenumber,page10of15